The computer age is over. Bandwidth and now “storewidth”, eclipse the PC paradigm. PCs remain important but peripheral. After a cataclysmic global run of 30 years, the PC revolution has stiffened into an establishment. So swiftly and subliminally did this silicon tide pass through the economy that many analysts missed much of the motion until it stopped. Then they mistook its dotage for dynamism. It congealed so fast that its Mount Rushmore giants still walk and talk.
Contemplate the monumental frieze looming over the road ahead, which ends at a revolutionary shrine: Bill Gates of Microsoft, Steven Jobs of Apple, Gordon Moore and Andrew Grove of Intel. Still very much around, they beam from magazine covers, iconize companies, orate at Davos and Comdex and publish books (Grove even writes his himself). Jobs launches new products in “insanely great” new tints and hues. Their totems tower over their time: Windows 2000, the millennial operating system launched with 30 million lines of code and a record-setting wild guess of 200,000 bugs; serried ranks of Pentium processors and support chips with scores of millions of transistors and instantly classical names, draining some 80 watts of power, enough to heat an igloo.
These Rushmore men, quick or dead, no longer shape the future. Computers no longer spearhead the economy. The action is elsewhere. The action is in the Telecosm.
Of course, the computer and the microchip remain enormously potent technologies. (So do the wheel, the steel mill and the nuclear power plant). Moore’s law, which dictates a doubling of microchip performance, or a halving of its cost, every 18 to 24 months, is still in force. The process of ingraining intelligence into every aspect of our lives, mind into every machine, tool or toy, continues at an accelerating pace. The displacement of matter by mind in the economy, already the most powerful economic event in recorded history — just now appearing in economic data — has not yet begun to end.
The computer era — the age of the microchip, which I term the Microcosm — is ending not because it has failed or even because it has been fulfilled but because the Microcosm itself has given birth to a new era. It has enabled a new technology that is transforming culture, economics and politics far more thoroughly than the computer age did.
The computer era is falling before the one technological force that could surpass in impact the computer’s ability to process information. That is communication, which is more essential to our humanity. Communication is the way we weave together a personality, a family, a business, a nation and a world. The Telecosm will make human communication universal, instantaneous, unlimited in capacity and free at the margins.
Because the computer made the creation and manipulation of information the central activity of the economy, this era has also been known as the information age. The great frustration of the computer era, however, has been the difficulty of communicating the information that we are told has become our most precious resource. Information is power, but information that cannot be readily moved is gridlock on the World Wide Wait. Immobile information makes our businesses larger, more static and hierarchical than they need to be. It makes our economies less flexible, our jobs less fulfilling, our lives less luminous with opportunity.
The Telecosm launches us beyond the copper cages of existing communications. The central event in technology over the last decade is a growing awareness that the information-bearing power of the electromagnetic spectrum — its bandwidth, that is, the range of frequencies and wavelengths available to carry signals — is not severely limited, as previously believed, but essentially infinite. From AM radio signals through microwaves to visible light — a band of frequencies millions of times the bandwidth of our 56 kilobit modems — the spectrum can carry usable signals.
An infinitude of potential bandwidth implies the endless multiplication of spectrum use and reuse. Such cellular technologies as CDMA allow the reuse of all available bandwidth in every cell, the sharing of cellular bandwidth among many users and the proliferation of local cells through the deployment of more antennas. The rise of Qualcomm, whose stock rose 27-fold in 1999, (see FORBES GLOBAL), and other related (GLOBAL story ), is based on this potential of ubiquitous waves.
The most powerful of all spectrum reuse technology is fiber optics. Every fiber-optic thread, only as thick as a human hair, can carry a thousand times more information on one path than all current wireless technologies put together. The basic measure of bandwidth is hertz, or wave cycles per second. The bandwidth of currently used wireless spectrum, running from AM radio to direct broadcast satellite, comes to a total of some 25 billion hertz (25 gigahertz), in scientific notation 25 x 10. The capacity of a fiber-optic cable — incorporating some 864 individual fibers — is measured in petahertz — 10 waves per second. Petahertz signifies a million gigahertz.
These technologies make up the Telecosm. It makes bandwidth — information at enormous speed and almost infinite scale — the defining abundance of a new era, eclipsing even the still fantastic abundance of the computer age.
Every new economic era has a defining abundance, a critical resource or technology that is expanding in production and plummeting in price so rapidly that it appears virtually free compared with an array of competing critical resources for which it can be substituted. These abundances come to define the very character of their age, whether an age of steam or oil or an age of information. Nations, companies, and individuals that exploit this “free” abundance gain market share against all rivals. The political leaders who accommodate them become the prime movers in global affairs. Their countries pioneer and prosper.
During the preindustrial era in America, the scarcity was horsepower, the abundance was land. In the Industrial Age, horsepower — physical force, translated eventually into watts or kilowatt-hours — abounded while land grew relatively scarce. Between 1660 and 1950, the cost of an effective kilowatt-hour dropped from thousands of dollars to some seven cents. The U.S. splurged on cheap horsepower — to clear farmland, to refine ores, to produce new farm equipment, to manufacture goods and capital gear and to contrive armaments for capturing or defending land. And these activities defined the industrial paradigm.
Over the last 30 years, the era of the Microcosm, transistors became asymptotically costless. On a computer memory chip the price of a transistor, with support circuits, dropped from some $7 to a few millionths of a cent. An MIPS (millions of instructions per second) of computer power that cost several million dollars in 1960 sells for less than a dollar today. Thirty-five years ago, a chip factory could produce a few score transistors a day. Today, a single production line in a microchip wafer fabrication facility (“fab”) can produce some 1.6 trillion transistors in 24 hours. (We end up “wasting” billions of them playing solitaire or singing karaoke, brushing our teeth, warming a sandwich or strumming “guitars” on music synthesizers.)
An era’s defining abundances relieve its critical scarcities. But abundances can also create new scarcities. The plethora of cheap fuel led to crowded roads and the need for pollution controls. The more recent glut of transistors — and the colossal streams of bits they shaped and sent — led to a shortage of the very communications capacity it was meant to enhance. While existing telephone bandwidth was ample for voice communications, it became suddenly scarce when faced with a global abundance of computers generating data at a rate of megabits per second. The canonical abundance of one era creates a canonical shortage for the next.
And now, in the Telecosm, all the defining abundances of the computer era — ever-cheaper power, transistors and silicon area — are becoming relatively scarce. And the crucial scarcity for which the transistor was meant to compensate — bandwidth — is the defining abundance of the new era. A global economy designed to waste transistors, power and silicon area — and conserve bandwidth above all — is breaking apart and reorganizing itself to waste bandwidth and conserve power, silicon area and transistors. The Microcosm has been turned upside down.
In the next era the most common digital devices will be cell phones and smart cards — increasingly performing an array of computer functions — that must be powered by low-power batteries or solar cells, technologies that double their efficiency over a span of decades rather than months.
Also scarce in these portable devices will be silicon area. Engineers must cram an ever-increasing array of radio frequency, digital-signal-processor and microprocessor functions within the constricted cavity and power budget of a handset. No longer can chips sprawl by the hundreds across PC backplanes and peripheral boards; cell phone computers will be based on single-chip systems like those from National Semiconductor, LSI Logic and Atmel, sharply economizing on power and silicon. Xilinx’s programmable logic devices, now meeting or exceeding the capabilities of application-specific chips and supreme in the rapidity and flexibility with which they can be programmed for specific applications, are crucial conservers of silicon area.
Similarly, coupling lasers and other communications devices to a fiber-optic thread with a 0.8 micron core, only a tenth as thick as a human hair, imposes severe restriction of space and power. These constraints become more acute when the fiber lies on the floor of the ocean, under enormous pressure. Led by Global Crossing, companies that achieve prowess in this hostile environment will present rivals with major barriers to entry.
Fueled by solar energy, satellites also restrict power and silicon area. Globalstar’s decision to use CDMA and keep most of the power-hungry electronics on the ground sealed its superiority over Iridium.
In the new millennium, the plummeting price of digital transistors will no longer spearhead the technology of the time. Crucial in the Telecosm, however, will be high-frequency analog devices linking digital appliances to the real world of sounds and pressures, images and movements, and user-interface inputs. Increasingly these analog digital converters, oscillators, low-noise amplifiers and high-frequency transmitters will employ exotic materials such as silicon germanium, gallium arsenide and indium phosphide in ever more complex combinations, called heterojunctions and high-electron mobility transistors. These transistors are millions of times more costly than the sturdy digital devices now being deployed by the billions on a single dynamic random-access memory (DRAM) chip for your personal computer.
Looking ahead, I will be focusing increasingly on low-power technologies, whether silicon germanium microchips from Atmel and AMCC, mixed signal chips from National Semiconductor, Texas Instruments and Analog Devices, or the low-power phones that made Qualcomm the leading stock of 1999 and will make it a spearhead communications company for the millennium. The critical analog function of a CDMA wireless phone, for instance, is the management of extremely low-power signals to keep multiple-shared-frequency transmissions from drowning each other out. The leading manufacturer of the power amps that do this job is Conexant, a reminder to investors that there are other CDMA plays available for those experiencing vertigo on the Qualcomm heights.